Bug #13970 Restart cluster w/ NoOfFragmentLogFiles set to low causes file system inconsist

• Submitted: 12 Oct 2005 17:32
• Modified: 24 Apr 2006 22:22
• Reporter: Jonathan Miller
• Status: Can't repeat
• Category: MySQL Cluster: Cluster (NDB) storage engine
• Severity: S2 (Serious)
• Version: 4.1->
• OS: Linux (Linux)
• Assigned to: Jonathan Miller
• CPU Architecture: Any

[12 Oct 2005 17:32] Jonathan Miller

Description:
After trying NoOfFragmentLogFiles set to 2, I set it back to orginal setting and tried to restart the cluster. Recieved:

Time: Wednesday 12 October 2005 - 18:19:04
Status: Ndbd file system error, restart node initial
Message: File not found (Ndbd file system inconsistency error, please report a bug)
Error: 2815
Error data: DBLQH: File system open failed. OS errno: 2
Error object: DBLQH (Line: 1824) 0x0000000a
Program: /home/ndbdev/jmiller/builds/libexec/ndbd
Pid: 5725
Trace: /space/run/ndb_6_trace.log.1
Version: Version 5.1.2 (a_drop5p5)
***EOM***

Current byte-offset of file-pointer is: 568

Time: Wednesday 12 October 2005 - 18:19:04
Status: Ndbd file system error, restart node initial
Message: File not found (Ndbd file system inconsistency error, please report a bug)
Error: 2815
Error data: DBLQH: File system open failed. OS errno: 2
Error object: DBLQH (Line: 1824) 0x0000000a
Program: /home/ndbdev/jmiller/builds/libexec/ndbd
Pid: 5728
Trace: /space/run/ndb_9_trace.log.1
Version: Version 5.1.2 (a_drop5p5)
***EOM***

Time: Wednesday 12 October 2005 - 18:19:04
Status: Ndbd file system error, restart node initial
Message: File not found (Ndbd file system inconsistency error, please report a bug)
Error: 2815
Error data: DBLQH: File system open failed. OS errno: 2
Error object: DBLQH (Line: 1824) 0x0000000a
Program: /home/ndbdev/jmiller/builds/libexec/ndbd
Pid: 15224
Trace: /space/run/ndb_5_trace.log.1
Version: Version 5.1.2 (a_drop5p5)
***EOM***

Current byte-offset of file-pointer is: 568

Time: Wednesday 12 October 2005 - 18:19:04
Status: Ndbd file system error, restart node initial
Message: File not found (Ndbd file system inconsistency error, please report a bug)
Error: 2815
Error data: DBLQH: File system open failed. OS errno: 2
Error object: DBLQH (Line: 1824) 0x0000000a
Program: /home/ndbdev/jmiller/builds/libexec/ndbd
Pid: 15227
Trace: /space/run/ndb_8_trace.log.1
Version: Version 5.1.2 (a_drop5p5)
***EOM***
--------------- Signal ----------------
r.bn: 247 "DBLQH", r.proc: 8, r.sigId: 907682 gsn: 260 "FSOPENREF" prio: 1
s.bn: 253 "NDBFS", s.proc: 8, s.sigId: 907680 length: 4 trace: 0 #sec: 0 fragInf: 0
 UserPointer: 767
 ErrorCode: 2815, File not found
 OS ErrorCode: 2
--------------- Signal ----------------
r.bn: 247 "DBLQH", r.proc: 5, r.sigId: 908342 gsn: 260 "FSOPENREF" prio: 1
s.bn: 253 "NDBFS", s.proc: 5, s.sigId: 908341 length: 4 trace: 0 #sec: 0 fragInf: 0
 UserPointer: 163
 ErrorCode: 2815, File not found
 OS ErrorCode: 2
--------------- Signal ----------------
r.bn: 247 "DBLQH", r.proc: 9, r.sigId: 905993 gsn: 260 "FSOPENREF" prio: 1
s.bn: 253 "NDBFS", s.proc: 9, s.sigId: 905992 length: 4 trace: 0 #sec: 0 fragInf: 0  UserPointer: 167
 ErrorCode: 2815, File not found
 OS ErrorCode: 2
--------------- Signal ----------------
r.bn: 247 "DBLQH", r.proc: 6, r.sigId: 906330 gsn: 260 "FSOPENREF" prio: 1
s.bn: 253 "NDBFS", s.proc: 6, s.sigId: 906329 length: 4 trace: 0 #sec: 0 fragInf: 0
 UserPointer: 367
 ErrorCode: 2815, File not found
 OS ErrorCode: 2

How to repeat:
see above

[18 Oct 2005 12:29] Jonathan Miller

Update

[20 Oct 2005 4:38] Tomas Ulin

setting lower prio

[6 Feb 2006 10:47] Jonas Oreland

I tried this wo/ sucess.
Can you try again?

[24 Apr 2006 13:52] Jonathan Miller

I have not been able to repeate. SK is currently testinf config.ini setting and will reopen if he hits it.

[25 Apr 2006 12:12] Serge Kozlov

I could to reproduce that but I'm not sure that it's a bug.
The bug appears if remove anything from DataDir/FileSystemPath directories (usually they are same) when ndbd nodes are starting.